CNN_Model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9353ad53",
   "metadata": {},
   "source": [
    "# Convolutional neural network for the human activities recognition\n",
    "* using preprocessed data in shape (samples, window-shape, components) to train and test the model\n",
    "* analyzes motor activities based on signals from sensors\n",
    "* its six activities: walking, standing, sitting, laying, walking up and downstairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ed065c2c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing necessary libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import glob\n",
    "import seaborn as sns\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras, lite\n",
    "from keras.models import Sequential\n",
    "from keras.layers.convolutional import Conv1D, MaxPooling1D\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4b5b3ac",
   "metadata": {},
   "source": [
    "# Loading and reshaping data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8aaa9e3d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading from the .txt files\n",
    "X_train = np.loadtxt('GeneralData/X_train.txt')\n",
    "X_test = np.loadtxt('GeneralData/X_test.txt')\n",
    "y_train = np.loadtxt('GeneralData/y_train.txt')\n",
    "y_test = np.loadtxt('GeneralData/y_test.txt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ef52abdc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((7279, 1152), (7279,))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of X and y data before reshaping\n",
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7feb1c85",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshaping y data\n",
    "y_test = np.reshape(y_test,(y_test.shape[0],1))\n",
    "y_train = np.reshape(y_train,(y_train.shape[0],1))\n",
    "# Changing the type (int) and decrease the indexes by one - to fit it with binary class matrix (output)\n",
    "y_test = y_test.astype(int) - 1\n",
    "y_train = y_train.astype(int) - 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c6b83ca1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Converts an y vectors to binary class matrix\n",
    "y_test_cat = keras.utils.to_categorical(y_test)\n",
    "y_train_cat = keras.utils.to_categorical(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "099bb1ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 1., 0., 0., 0., 0.],\n",
       "       [0., 0., 1., 0., 0., 0.],\n",
       "       [0., 0., 0., 1., 0., 0.],\n",
       "       [0., 0., 0., 0., 1., 0.],\n",
       "       [0., 1., 0., 0., 0., 0.]], dtype=float32)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Visualisating output\n",
    "y_test_cat[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4969d3ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshaping X data into (samples, windows-shape, compontents) \n",
    "# values from the SingalPreprocessing\n",
    "samples = 10399\n",
    "window_shape = 128\n",
    "components = 9\n",
    "\n",
    "X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1] // components, components))\n",
    "X_train = np.reshape (X_train, (X_train.shape[0], X_train.shape[1] // components, components))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b0d0f704",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((7279, 128, 9), (3120, 128, 9))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# X after reshaping\n",
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce54e241",
   "metadata": {},
   "source": [
    "* Data is prepared to use into training process"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e6a9162",
   "metadata": {},
   "source": [
    "# Modeling CNN\n",
    "* fitting\n",
    "* testing and evaluating"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "47a3d4df",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating model\n",
    "def createModel(nr_filter, nr_kernel):\n",
    "    model = Sequential()\n",
    "    # Adding layers\n",
    "    # first convolution\n",
    "    model.add(Conv1D(filters=nr_filter, kernel_size=nr_kernel, input_shape=X_train.shape[1:], activation='relu'))\n",
    "    # second convolution, the number of filteres divided by 2\n",
    "    model.add(Conv1D(nr_filter/2, nr_kernel, activation='relu'))\n",
    "    # max pooling size 2\n",
    "    model.add(MaxPooling1D(2))\n",
    "    # only 50% of data goes on\n",
    "    model.add(Dropout(0.5))\n",
    "    # flattering the matrix\n",
    "    model.add(Flatten())\n",
    "    # deeply connected layer with 125 units\n",
    "    model.add(Dense(units=125, activation='relu'))\n",
    "    # deeply connected layer with 6 units = 6 activities\n",
    "    model.add(Dense(units=6, activation='softmax'))\n",
    "    # Model Visualisation\n",
    "    model.summary()\n",
    "    # Model compilation\n",
    "    model.compile(optimizer='adam', loss = 'categorical_crossentropy', metrics=['accuracy'])\n",
    "    # Return model\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a2cbf7eb",
   "metadata": {},
   "source": [
    "* the optimizer is adam\n",
    "* using categorical cross entropy because labels are one-hot represenation and more then one class\n",
    "* two dropout layers to avoid overfitting\n",
    "* activation function is ReLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "881fce39",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Getting model with properties\n",
    "def getModel(epochs, batch, nr_filter, nr_kernel):\n",
    "    # Model creating\n",
    "    model = createModel(nr_filter,nr_kernel)\n",
    "    # Model fitting\n",
    "    model.fit(x=X_train, y=y_train_cat, epochs=epochs, batch_size=batch, validation_data=(X_test,y_test_cat))\n",
    "    # Getting model properties\n",
    "    model_feat = pd.DataFrame(model.history.history)\n",
    "    # Getting loss values\n",
    "    model_loss = model_feat[['loss','val_loss']]\n",
    "    # Return model, model loss and model properties\n",
    "    return model, model_loss, model_feat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "2ed94e79",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "\n",
    "# Model properties\n",
    "def propModel(model, loss):\n",
    "    # predictions of model using testing dataset\n",
    "    y_pred = model.predict(X_test)\n",
    "    # changing form of output data\n",
    "    y_pred = np.argmax(y_pred, axis=1)\n",
    "    y_real = np.argmax(y_test_cat, axis=1)\n",
    "    # showing properties\n",
    "    print('--------------------------------Classification report-')\n",
    "    print(' ')\n",
    "    print(classification_report(y_real,y_pred))\n",
    "    print(' ')\n",
    "    print('-------------------------------------Confusion matrix-')\n",
    "    print(' ')\n",
    "    print(confusion_matrix(y_real,y_pred))\n",
    "    print(' ')\n",
    "    print('-------------------------------------Loss Values plot-')\n",
    "    loss.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "91ea5180",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fitting components\n",
    "\n",
    "epochs_nr = 11 # number of epochs during the training, it is a right value when i checked a model\n",
    "batch_val = 29 # batch size 7279/29 = 251"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "675b1ebb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_14\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_28 (Conv1D)          (None, 124, 128)          5888      \n",
      "                                                                 \n",
      " conv1d_29 (Conv1D)          (None, 120, 64)           41024     \n",
      "                                                                 \n",
      " max_pooling1d_14 (MaxPoolin  (None, 60, 64)           0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_22 (Dropout)        (None, 60, 64)            0         \n",
      "                                                                 \n",
      " flatten_14 (Flatten)        (None, 3840)              0         \n",
      "                                                                 \n",
      " dense_28 (Dense)            (None, 125)               480125    \n",
      "                                                                 \n",
      " dense_29 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 527,793\n",
      "Trainable params: 527,793\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 9s 32ms/step - loss: 0.3569 - accuracy: 0.8557 - val_loss: 0.2099 - val_accuracy: 0.9234\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 9s 36ms/step - loss: 0.1592 - accuracy: 0.9371 - val_loss: 0.1694 - val_accuracy: 0.9295\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 9s 37ms/step - loss: 0.1327 - accuracy: 0.9444 - val_loss: 0.1288 - val_accuracy: 0.9471\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 9s 38ms/step - loss: 0.1315 - accuracy: 0.9464 - val_loss: 0.1355 - val_accuracy: 0.9442\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 10s 38ms/step - loss: 0.1351 - accuracy: 0.9494 - val_loss: 0.1217 - val_accuracy: 0.9494\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 10s 39ms/step - loss: 0.1219 - accuracy: 0.9496 - val_loss: 0.1176 - val_accuracy: 0.9490\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 10s 39ms/step - loss: 0.1222 - accuracy: 0.9494 - val_loss: 0.1354 - val_accuracy: 0.9429\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 10s 39ms/step - loss: 0.1111 - accuracy: 0.9545 - val_loss: 0.1219 - val_accuracy: 0.9455\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 10s 39ms/step - loss: 0.1059 - accuracy: 0.9562 - val_loss: 0.1196 - val_accuracy: 0.9494\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 10s 40ms/step - loss: 0.1093 - accuracy: 0.9545 - val_loss: 0.1317 - val_accuracy: 0.9433\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 10s 41ms/step - loss: 0.1159 - accuracy: 0.9515 - val_loss: 0.1582 - val_accuracy: 0.9417\n"
     ]
    }
   ],
   "source": [
    "# Model: filter size = 128, kernel size = 5\n",
    "\n",
    "model1, loss1, prop1 = getModel(epochs_nr,batch_val, 128, 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "959feeb9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 11ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      1.00      0.99       532\n",
      "           1       0.97      1.00      0.98       462\n",
      "           2       1.00      0.94      0.97       421\n",
      "           3       0.87      0.84      0.86       530\n",
      "           4       0.86      0.88      0.87       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.94      3120\n",
      "   macro avg       0.94      0.94      0.94      3120\n",
      "weighted avg       0.94      0.94      0.94      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[530   0   0   0   2   0]\n",
      " [  2 460   0   0   0   0]\n",
      " [ 10  16 395   0   0   0]\n",
      " [  0   0   0 446  84   0]\n",
      " [  1   0   0  67 518   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model1,loss1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "05f54831",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_15\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_30 (Conv1D)          (None, 126, 128)          3584      \n",
      "                                                                 \n",
      " conv1d_31 (Conv1D)          (None, 124, 64)           24640     \n",
      "                                                                 \n",
      " max_pooling1d_15 (MaxPoolin  (None, 62, 64)           0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_23 (Dropout)        (None, 62, 64)            0         \n",
      "                                                                 \n",
      " flatten_15 (Flatten)        (None, 3968)              0         \n",
      "                                                                 \n",
      " dense_30 (Dense)            (None, 125)               496125    \n",
      "                                                                 \n",
      " dense_31 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 525,105\n",
      "Trainable params: 525,105\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 8s 30ms/step - loss: 0.3730 - accuracy: 0.8511 - val_loss: 0.1869 - val_accuracy: 0.9260\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 7s 29ms/step - loss: 0.1743 - accuracy: 0.9338 - val_loss: 0.1593 - val_accuracy: 0.9288\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 8s 30ms/step - loss: 0.1521 - accuracy: 0.9397 - val_loss: 0.1537 - val_accuracy: 0.9407\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 8s 31ms/step - loss: 0.1307 - accuracy: 0.9461 - val_loss: 0.1334 - val_accuracy: 0.9458\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 8s 32ms/step - loss: 0.1261 - accuracy: 0.9483 - val_loss: 0.1444 - val_accuracy: 0.9446\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 8s 31ms/step - loss: 0.1174 - accuracy: 0.9500 - val_loss: 0.1775 - val_accuracy: 0.9285\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 8s 32ms/step - loss: 0.1205 - accuracy: 0.9493 - val_loss: 0.1240 - val_accuracy: 0.9449\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 8s 33ms/step - loss: 0.1158 - accuracy: 0.9518 - val_loss: 0.1257 - val_accuracy: 0.9462\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 8s 32ms/step - loss: 0.1197 - accuracy: 0.9519 - val_loss: 0.1319 - val_accuracy: 0.9468\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 8s 32ms/step - loss: 0.1077 - accuracy: 0.9556 - val_loss: 0.1180 - val_accuracy: 0.9474\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 8s 33ms/step - loss: 0.1031 - accuracy: 0.9559 - val_loss: 0.1209 - val_accuracy: 0.9487\n"
     ]
    }
   ],
   "source": [
    "# Second model: filter size = 128, kernel size = 3\n",
    "\n",
    "model2, loss2, prop2 = getModel(epochs_nr,batch_val, 128, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "c3a89ebb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 5ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.98      0.99       532\n",
      "           1       0.98      1.00      0.99       462\n",
      "           2       0.99      1.00      0.99       421\n",
      "           3       0.86      0.87      0.86       530\n",
      "           4       0.88      0.86      0.87       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.95      3120\n",
      "   macro avg       0.95      0.95      0.95      3120\n",
      "weighted avg       0.95      0.95      0.95      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[523   5   4   0   0   0]\n",
      " [  1 461   0   0   0   0]\n",
      " [  0   2 419   0   0   0]\n",
      " [  1   1   0 462  66   0]\n",
      " [  2   1   0  77 506   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model2,loss2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "76439f54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_16\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_32 (Conv1D)          (None, 122, 128)          8192      \n",
      "                                                                 \n",
      " conv1d_33 (Conv1D)          (None, 116, 64)           57408     \n",
      "                                                                 \n",
      " max_pooling1d_16 (MaxPoolin  (None, 58, 64)           0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_24 (Dropout)        (None, 58, 64)            0         \n",
      "                                                                 \n",
      " flatten_16 (Flatten)        (None, 3712)              0         \n",
      "                                                                 \n",
      " dense_32 (Dense)            (None, 125)               464125    \n",
      "                                                                 \n",
      " dense_33 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 530,481\n",
      "Trainable params: 530,481\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 11s 43ms/step - loss: 0.3531 - accuracy: 0.8597 - val_loss: 0.1827 - val_accuracy: 0.9250\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 12s 47ms/step - loss: 0.1509 - accuracy: 0.9404 - val_loss: 0.1583 - val_accuracy: 0.9349\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 12s 47ms/step - loss: 0.1392 - accuracy: 0.9426 - val_loss: 0.1312 - val_accuracy: 0.9458\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 12s 47ms/step - loss: 0.1251 - accuracy: 0.9501 - val_loss: 0.1325 - val_accuracy: 0.9497\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 12s 49ms/step - loss: 0.1189 - accuracy: 0.9511 - val_loss: 0.1431 - val_accuracy: 0.9429\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 13s 52ms/step - loss: 0.1144 - accuracy: 0.9530 - val_loss: 0.1151 - val_accuracy: 0.9510\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 12s 50ms/step - loss: 0.1163 - accuracy: 0.9534 - val_loss: 0.1410 - val_accuracy: 0.9417\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 13s 51ms/step - loss: 0.1213 - accuracy: 0.9504 - val_loss: 0.1516 - val_accuracy: 0.9449\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 13s 53ms/step - loss: 0.1403 - accuracy: 0.9459 - val_loss: 0.1312 - val_accuracy: 0.9465\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 14s 56ms/step - loss: 0.1090 - accuracy: 0.9545 - val_loss: 0.1183 - val_accuracy: 0.9506\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 16s 66ms/step - loss: 0.1060 - accuracy: 0.9565 - val_loss: 0.1104 - val_accuracy: 0.9503\n"
     ]
    }
   ],
   "source": [
    "# Next model: filter size = 128, kernel size = 7\n",
    "\n",
    "model3, loss3, prop3 = getModel(epochs_nr,batch_val, 128, 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "a0e13a1e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 11ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       532\n",
      "           1       1.00      1.00      1.00       462\n",
      "           2       1.00      1.00      1.00       421\n",
      "           3       0.82      0.92      0.87       530\n",
      "           4       0.92      0.81      0.86       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.95      3120\n",
      "   macro avg       0.95      0.95      0.95      3120\n",
      "weighted avg       0.95      0.95      0.95      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[532   0   0   0   0   0]\n",
      " [  1 461   0   0   0   0]\n",
      " [  0   1 420   0   0   0]\n",
      " [  0   0   0 486  44   0]\n",
      " [  1   1   0 107 477   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD4CAYAAADiry33AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAtwklEQVR4nO3deXxU1f3/8dcn+zoQQsjKEiCCCVGQgFoLWkFcWgHBFteqVXGv1tZqrVZrtbXab21/LW61aGv1q1SxxYoiFRT9KsouhjVEliRAErJA9mTm/P64EzKEQCZkksnc+Twfjzwyc+femTMs7zlzzrmfK8YYlFJK2VeIvxuglFKqZ2nQK6WUzWnQK6WUzWnQK6WUzWnQK6WUzYX5uwHtDRw40AwbNszfzVBKqYCyZs2acmNMUkeP9bmgHzZsGKtXr/Z3M5RSKqCIyK5jPaZDN0opZXMa9EopZXMa9EopZXN9boxeKRWcmpubKSoqoqGhwd9N6dOioqLIyMggPDzc62M06JVSfUJRURHx8fEMGzYMEfF3c/okYwwHDhygqKiIzMxMr4/ToRulVJ/Q0NBAYmKihvxxiAiJiYld/tajQa+U6jM05Dt3In9Gtgn66rpm/vjf7WwsqvZ3U5RSqk/xKuhF5AIR2SoiBSJyXweP3ywiG0VkvYh8IiLZ7u3DRKTevX29iDzr6zfQKiQEnvrvNlZsL+upl1BK2VxcXJy/m9AjOp2MFZFQYB5wHlAErBKRRcaYTR67vWqMeda9/3Tg98AF7sd2GGPG+rTVHYiPCmdoYgybSg729EsppVRA8aZHPxEoMMYUGmOagNeAGZ47GGM80zUW8Mtlq7JTHeSX6NCNUqp7jDHcc889jBkzhtzcXF5//XUA9u7dy+TJkxk7dixjxozh448/xul0cu211x7e96mnnvJz64/mzfLKdGCPx/0i4PT2O4nIbcDdQARwrsdDmSKyDjgIPGCM+biDY+cCcwGGDBnidePby0518O5X+6hpbCEuUleOKhWofvl2vs+/nWenOXjo4hyv9l24cCHr169nw4YNlJeXM2HCBCZPnsyrr77K+eefz89//nOcTid1dXWsX7+e4uJivvrqKwCqqqp82m5f8NlkrDFmnjFmBHAv8IB7815giDFmHNaHwKsi4ujg2OeNMXnGmLykpA6Lr3klO8166i17dfhGKXXiPvnkEy6//HJCQ0NJTk7m7LPPZtWqVUyYMIEXX3yRhx9+mI0bNxIfH8/w4cMpLCzkjjvu4L333sPhOCri/M6bbm8xMNjjfoZ727G8BjwDYIxpBBrdt9eIyA7gJKBHylPmpPUDYNPeg+QNG9ATL6GU6gXe9rx72+TJk1mxYgXvvPMO1157LXfffTff//732bBhA0uWLOHZZ59lwYIFzJ8/399NPYI3PfpVQJaIZIpIBHAZsMhzBxHJ8rj7bWC7e3uSezIXERkOZAGFvmh4R5IdkQyIjdAJWaVUt0yaNInXX38dp9NJWVkZK1asYOLEiezatYvk5GRuvPFGbrjhBtauXUt5eTkul4vZs2fz6KOPsnbtWn83/yid9uiNMS0icjuwBAgF5htj8kXkEWC1MWYRcLuITAWagUrgGvfhk4FHRKQZcAE3G2MqeuKNgHUiQXaqg006dKOU6oZLLrmEzz77jFNPPRUR4YknniAlJYW//e1vPPnkk4SHhxMXF8ff//53iouLue6663C5XAD85je/8XPrjybG+GWBzDHl5eWZ7lx45NeLN/PSpzvJ/+X5hIfa5nwwpWxv8+bNnHzyyf5uRkDo6M9KRNYYY/I62t92SZid6qCpxUVhWa2/m6KUUn2C/YLevfJm015dT6+UUmDDoB8+MJbIsBCdkFVKKTfbBX1YaAijU+J1QlYppdxsF/RgDd9sKjlIX5toVkopf7Bn0Kc6qKxrZm+1XpJMKaXsGfStE7I6Tq+UUvYM+tEpDkTQcXqlVI85Xu36nTt3MmbMmF5szfHZMuhjI8PITIzVHr1SSuFdUbOAdHKaQy8rqFSgevc+2LfRt8+ZkgsXPn7Mh++77z4GDx7MbbfdBsDDDz9MWFgYy5cvp7KykubmZh599FFmzJhxzOfoSENDA7fccgurV68mLCyM3//+93zrW98iPz+f6667jqamJlwuF2+++SZpaWl873vfo6ioCKfTyYMPPsicOXO69bbBxkGfnergnS/3crChGUdUuL+bo5Tq4+bMmcNdd911OOgXLFjAkiVL+OEPf4jD4aC8vJwzzjiD6dOnd+kC3fPmzUNE2LhxI1u2bGHatGls27aNZ599ljvvvJMrr7ySpqYmnE4nixcvJi0tjXfeeQeA6mrfdFbtG/TuCdnNJQc5fXiin1ujlOqS4/S8e8q4ceMoLS2lpKSEsrIyEhISSElJ4Uc/+hErVqwgJCSE4uJi9u/fT0pKitfP+8knn3DHHXcAMHr0aIYOHcq2bds488wzeeyxxygqKmLWrFlkZWWRm5vLj3/8Y+69916+853vMGnSJJ+8N1uO0QPkpLaWQtBxeqWUd7773e/yxhtv8PrrrzNnzhxeeeUVysrKWLNmDevXryc5OZmGBt8s277iiitYtGgR0dHRXHTRRSxbtoyTTjqJtWvXkpubywMPPMAjjzzik9eybY8+KT6SgXFam14p5b05c+Zw4403Ul5ezkcffcSCBQsYNGgQ4eHhLF++nF27dnX5OSdNmsQrr7zCueeey7Zt29i9ezejRo2isLCQ4cOH88Mf/pDdu3fz5ZdfMnr0aAYMGMBVV11F//79eeGFF3zyvmwb9CJCdlo/7dErpbyWk5PDoUOHSE9PJzU1lSuvvJKLL76Y3Nxc8vLyGD16dJef89Zbb+WWW24hNzeXsLAwXnrpJSIjI1mwYAEvv/wy4eHhpKSkcP/997Nq1SruueceQkJCCA8P55lnnvHJ+7JdPXpPj7+7hfmffM1XvzyfiDDbjlIpZQtaj957QV+P3lN2moMmp4uC0hp/N0UppfzGtkM3YC2xBGtCtnUVjlJK+crGjRu5+uqrj9gWGRnJ559/7qcWdczWQZ85MJbo8FBrQna8v1ujlOqMMaZLa9T9LTc3l/Xr1/fqa57IcLuth25CQ4TRqfF6tSmlAkBUVBQHDhzQ8uLHYYzhwIEDREVFdek4W/fowRq+eXtDScD1FJQKNhkZGRQVFVFWVubvpvRpUVFRZGRkdOkY+wd9moNXPt9NUWU9gwfE+Ls5SqljCA8PJzMz09/NsCVbD93AkROySikVjGwf9KNTHISIXoREKRW8bB/00RGhZA6M1R69Uipo2T7oAXLS+mmPXikVtIIi6LPTHBRX1VNd1+zvpiilVK8LjqB3T8jm63p6pVQQCoqgP7l15Y0O3yilglBQBH1SfCSD4iN1QlYpFZSCIujBGqfXHr1SKhh5FfQicoGIbBWRAhG5r4PHbxaRjSKyXkQ+EZFsj8d+5j5uq4ic78vGd0VOmoOC0hoaW5z+aoJSSvlFp0EvIqHAPOBCIBu43DPI3V41xuQaY8YCTwC/dx+bDVwG5AAXAE+7n6/XZaf2o8Vl2L5fa9MrpYKLNz36iUCBMabQGNMEvAbM8NzBGOM5JhILtJafmwG8ZoxpNMZ8DRS4n6/Xtdaj1+EbpVSw8aaoWTqwx+N+EXB6+51E5DbgbiACONfj2JXtjk3v4Ni5wFyAIUOGeNPuLhs6IIaYiFCdkFVKBR2fTcYaY+YZY0YA9wIPdPHY540xecaYvKSkJF816QghIcLJqTohq5QKPt4EfTEw2ON+hnvbsbwGzDzBY3tUTpqDTXsP4nLphQ2UUsHDm6BfBWSJSKaIRGBNri7y3EFEsjzufhvY7r69CLhMRCJFJBPIAr7ofrNPTHaqg5rGFooq6/3VBKWU6nWdjtEbY1pE5HZgCRAKzDfG5IvII8BqY8wi4HYRmQo0A5XANe5j80VkAbAJaAFuM8b4bX1j64Rsfkk1QxL1IiRKqeDg1RWmjDGLgcXttv3C4/adxzn2MeCxE22gL52UHE9oiLBp70EuzE31d3OUUqpXBM2ZsQBR4aGMSIrVCVmlVFAJqqAHa5xel1gqpYJJ0AV9Tlo/9lY3UFHb5O+mKKVUrwi6oG+dkN2svXqlVJAIuqBvrU2fX6IXIVFKBYegC/oBsRGk9ovSCVmlVNAIuqAHnZBVSgWX4Az6NAc7ymppaNba9Eop+wvKoM9Jc+B0GbbtP+TvpiilVI8LyqDPTu0HQL6O0yulgkBQBn1GQjTxkWE6IauUCgpBGfSHa9PrhKxSKggEZdCDNSG7WWvTK6WCQFAHfV2Tk10Vdf5uilJK9ajgDfpUvVi4Uio4BG3QZyXHERYiWgpBKWV7QRv0kWGhjBwUpxOySinbC9qgB2ucXodulFJ2F9xBn+qg9FAjZYca/d0UpZTqMUEd9Dlp1hmyWpteKWVnQR30h1feaNArpWwsqIO+X0w46f2jteaNUsrWgjrooXVCVpdYKqXsS4M+1UFheS11TS3+bopSSvUIDfo0B8bA1n1am14pZU9BH/Q5aTohq5Syt6AP+vT+0TiiwnRCVillW0Ef9CKiZ8gqpWwt6IMerEsLbtl3EKfWpldK2ZAGPdaEbEOzi6/La/3dFKWU8jkNevQMWaWUvXkV9CJygYhsFZECEbmvg8fvFpFNIvKliHwgIkM9HnOKyHr3zyJfNt5XRg6KIyI0RMfplVK2FNbZDiISCswDzgOKgFUissgYs8ljt3VAnjGmTkRuAZ4A5rgfqzfGjPVts30rIiyErOQ4vQiJUsqWvOnRTwQKjDGFxpgm4DVghucOxpjlxpjWi6+uBDJ828yel51qrbwxRidklVL24k3QpwN7PO4Xubcdy/XAux73o0RktYisFJGZHR0gInPd+6wuKyvzokm+l53m4EBtk9amV0rZjk8nY0XkKiAPeNJj81BjTB5wBfAHERnR/jhjzPPGmDxjTF5SUpIvm+S11gnZfJ2QVUrZjDdBXwwM9rif4d52BBGZCvwcmG6MOdwtNsYUu38XAh8C47rR3h5zcmspBJ2QVUrZjDdBvwrIEpFMEYkALgOOWD0jIuOA57BCvtRje4KIRLpvDwTOAjwncfsMR1Q4QwbE6BJLpZTtdLrqxhjTIiK3A0uAUGC+MSZfRB4BVhtjFmEN1cQB/xQRgN3GmOnAycBzIuLC+lB5vN1qnT6ldUJWKaXspNOgBzDGLAYWt9v2C4/bU49x3KdAbnca2Juy0xws2bSPmsYW4iK9+qNRSqk+T8+M9ZCd2lqbXnv1Sin70KD3kK0TskopG9Kg95DaL4qEmHCdkFVK2YoGvYfW2vR6ERKllJ1o0LeTnepgy75DtDhd/m6KUkr5hAZ9O9lpDppaXBRqbXqllE1o0LeTndoP0AlZpZR9aNC3MzwploiwEJ2QVUrZhgZ9O+GhIYxOidcevVLKNjToO5Cd6iC/pFpr0yulbEGDvgPZaQ4q65rZd7DB301RSqlu06DvwOGLhevwjVLKBjToOzA61YGIBr1Syh406DsQFxnGsMRYXXmjlLIFDfpjsCZkNeiVUoFPg/4YstMc7K6o42BDs7+bopRS3aJBfwytE7Jb9h7yc0uUUqp7NOiPoa02fbWfW6KUUt2jQX8Mg+IjSYyN0AlZpVTA06A/htba9Br0SqlAp0F/HNlpDrbtq6FZa9MrpQKYBv1xZKc6aHK6KCit8XdTlFLqhGnQH0eOXixcKWUDGvTHkTkwjqhwrU2vlApsGvTHERoijEpxaI9eKRXQNOg7keNeeaO16ZVSgUqDvhPZqQ6q65sprqr3d1OUUuqEaNB3IlsnZJVSAU6DvhOjU+Kt2vQ6IauUClAa9J2IiQgjc2Cs9uiVUgFLg94L2alaCkEpFbi8CnoRuUBEtopIgYjc18Hjd4vIJhH5UkQ+EJGhHo9dIyLb3T/X+LLxvSUnrR9FlfVU12lteqVU4Ok06EUkFJgHXAhkA5eLSHa73dYBecaYU4A3gCfcxw4AHgJOByYCD4lIgu+a3zsOT8hqr14pFYC86dFPBAqMMYXGmCbgNWCG5w7GmOXGmDr33ZVAhvv2+cBSY0yFMaYSWApc4Jum957Wi5Bo0CulApE3QZ8O7PG4X+TedizXA+925VgRmSsiq0VkdVlZmRdN6l1J8ZEkxUfqhKxSKiD5dDJWRK4C8oAnu3KcMeZ5Y0yeMSYvKSnJl03yGZ2QVUoFKm+CvhgY7HE/w73tCCIyFfg5MN0Y09iVYwNBTpqDgtJDNLVobXqlVGDxJuhXAVkikikiEcBlwCLPHURkHPAcVsiXejy0BJgmIgnuSdhp7m0BJzvNQbPTsG2/XixcKRVYOg16Y0wLcDtWQG8GFhhj8kXkERGZ7t7tSSAO+KeIrBeRRe5jK4BfYX1YrAIecW8LODohq5QKVGHe7GSMWQwsbrftFx63px7n2PnA/BNtYF8xNDGWmIhQnZBVSgUcPTPWS6EhwuiUeO3RK6UCjgZ9F2SnOdhcorXplVKBRYO+C3LS+nGosYU9FVqbXikVODTou6BtQrbazy1RSinvadB3waiUeEJEL0KilAosGvRdEBUeyoikOJ2QVUoFFHsF/a5PweXs0ZfITnNoj14pFVDsE/Tl2+Glb8Nfz4P9+T32MtmpDkqqG6isbeqx11BKKV+yT9AnjoRZf4HKnfDcZFj2GLQ0dnpYV+Wk9QP0DFmlVOCwT9CLQO6lcNsqGHMprHgCnv0m7F7p05c5OTUe0AlZpVTgsE/Qt4pNhFnPwZVvQnMDzL8A3vkJNPqmGFliXCQpjijt0SulAob9gr5V1lS49TM4/WZY9QLMOx22+aZwpk7IKqUCiX2DHiAyDi58HK5fCpEOePV78Mb1UNO9q1hlpzooKKuhoblnV/gopZQv2DvoWw2eADetgHPuh03/hnkTYMNrcII1a3LSHDhdhu37a3zcUKWU8r3gCHqAsAg45164+RMYeBK8dRP8YzZU7uryU2WnWaUQ8ku0FIJSqu8LnqBvNWg0XPceXPQ72PM5PH0mrHymSydaDU6IIS4yTCdklVIBIfiCHiAkBCbeCLeuhGFnwXv3wV+nwf5NXh4unJwarxOySqmAEJxB36r/YLhiAcz+K1R+bZ1otfzXXp1olZ3qYPPeg7hcWpteKdW3BXfQQ7sTrWbBR7+FZyfB7s+Pe1h2moPaJie7K+p6qaFKKXViNOhbxSbCrOfdJ1rVwfzzj3uiVWsphHwdvlFK9XEa9O1lTbXG7k+/6bgnWo0cFEdYiOhFSJRSfZ4GfUci4+DC3x59olVt+eFdosJDGTkoTidklVJ9ngb98bQ/0erPR55olZ3q0CWWSqk+T4O+M54nWiWObDvRqmo32WkO9h9spLzG9+WQlVLKVzTovTVoNPxgCVz4pHWi1bwzmFL9JiG42Ky9eqVUH6ZB3xUhIXD6XGuydug3yFz9KG9GPEzJtrX+bplSSh2TBv2J6D8YrvwnzHqBYSGlzF51BSz/DTib/d0ypZQ6igb9iRKBU77LL4e8xIfhZ8FHj1sXOako9HfLlFLqCBr03TRk8GDm1txE08y/woHt1lm161454RLISinlaxr03ZSd6sBlYFPiVLjlU0gdC/++Ff55DdRV+Lt5SimlQd9dOZ616ftlwDWLYOrDsOUdeOYs+HqFfxuolAp6XgW9iFwgIltFpEBE7uvg8ckislZEWkTk0naPOUVkvftnka8a3ldkJEQTHxXWdoZsSCh880dww38hIhb+Nh3efxBamvzbUKW6qrkedq+E+ip/t0R1U1hnO4hIKDAPOA8oAlaJyCJjjGfx9t3AtcBPOniKemPM2O43tW8SEbJTHSzfUsqCVXuYcvIgEuMiIW0c3PQRLPk5fPr/oPBDmP0CJI3yd5OV6pgxUL4NCv4LBR/Arv+DlgaI6g+TfgwT50J4lL9bqU5Ap0EPTAQKjDGFACLyGjADOBz0xpid7sdcPdDGPu/qM4fy63c289M3vyREYPzQBM7LTmZadgrDLv4DZJ0Hi+6A586G8x+DvB9Yq3aU8rf6Kvj6I3e4L4ODRdb2xCwYfy1kTIAN/wtLH4TPn4Vv3Q+nXm59c1UBQ0wnq0PcQzEXGGNucN+/GjjdGHN7B/u+BPzHGPOGx7YWYD3QAjxujPlXB8fNBeYCDBkyZPyuXV2/jqu/GWPILznI+5v2s3TT/sNny2YNimNaTjIXDhNyPr8PKVwGJ10I0/8EcUl+brUKOi4nlKyHHR9Y4V60GozTKt6XORlGToWRU6D/kCOP+3oFLH0IStZC0miY8gsYdZF2WPoQEVljjMnr8LFeCPp0Y0yxiAwHlgFTjDE7jvV6eXl5ZvXq1V68rb5tT0UdS92h/8XOCpwuQ3JcOA8O+piL9j2LRPdHZj5t9faV6kmH9llDMTs+gB3Lob4CEEgbCyOmWOGekQeh4cd/HmOs4n7LfgUHCmDw6TD1lzD0zN54F6oTxwt6b4ZuioHBHvcz3Nu8Yowpdv8uFJEPgXHAMYPeLgYPiOEH38zkB9/MpKquiWVbSlm6aT8/3XYWf2pO40/OeZz0yqUUDr+SxJmP08/h8HeTlV20NMLuz9zhvgz2f2Vtjx0EJ51vhfuIb0HswK49rwjkzITR34Z1/4APH4cXL7C+oU75BSRn+/ytKN/wpkcfBmwDpmAF/CrgCmNMfgf7voRHj15EEoA6Y0yjiAwEPgNmtJvIPYJdevTH0tDs5NMd5Sz/ajcn5z/FFeYdtrkyeDHlAUaPPZPzspNJ6x/t72aqQGKMdUZ2gXs4ZufH1lXSQsJhyBnWUMzIqZA8xrdDLU118Pkz8MkfofEgjL0CzvmZVSJE9bpuDd24n+Ai4A9AKDDfGPOYiDwCrDbGLBKRCcBbQALQAOwzxuSIyDeA5wAX1lLOPxhj/nq817J70HtyuQw7Vv6b1OV3E9FczW+bL2O+8wJy0vtz3skpTMtJZnRKPKLjoBZnMxStgspdkHNJcK8AaTxkjZu3hnuVe15rwHD3cMwUGDbJuohOT6urgI//B774i3V/4o3WKp2YAT3/2uqwbgd9bwqmoD+sttxalbN1MXsSzuDhkNtYVhKKMdY6/dYVPBOGJRAWGmTnuFV87Z44XGYFW5P7Gr7JufDdF2Fgln/b11tcLtj3ZdufxZ6V4GqBiDhrEnXEuVa4DxjuvzZW7YEPf2Ot0omIg7PuhDNusc4nUT1Ogz4QGANrXoT37ofwaKrO+x/ebclj6ab9fFJQTlOLi/4x4Zw7ehDTspOZfFISMRHeTLEEmMYaa+ihdfKwtUhcvyEw8lyrtwrw9p3WWPS3f2cNGdiVywkrn4b/+yPUllnbUk6xQn3EFGtCNCzCv21sr3QzfPAIbF0McSnWhXvGXd35ZK+HAzWNVNU3MyKpF76R2IQGfSAp2wYLb4C9G+C0a+CC31BrIlmxrYylm/bzwZZSquubiQgLYdLIgZyXncyFY1LpF+P9f6I+xeWC/RvbJg53rwRXM4THWEMPrYGWOOLI8eWDJbBwrvWhcMplVuBHxvvvffSEikL4163WxOrIqZD7XavnHjfI3y3zzq7P4L8PW98+EkfCuQ9C9ozjzhOUHWrk+RU7eHnlLlqchoen53DVGUN7r80BTIM+0LQ0wfLHrF5c4giY9RdIP816yOnii50Vh5duFlXWExEawtTsQcwal8HZo5II7+vDOzWlVqgXfACFy9t6qsm5bb32IWdAWOTxn8fltMaGP/wNJGTCpfOtJYOBzhhYPd8qnRESBhc9AafMCcw168bA1nfhg19C2RZIH2/VgsqcfMRungHf1OJi5rh0Kmqb+HBrGVefMZRfXJzd9/9d+5kGfaD6egW8dTPU7LfOSDzrriPOSDTG8FXxQRauK2LR+hIO1DYxMC6C6aemM+u0dHLSHH1jIrelyerVtQ7H7NtobY8ZaC3zGzHF6qnGJ5/Y8+/8P3jzBqgrh/N+BaffFJihCFBdDItutz4Ih38LZvzZKpYX6FxO2PCa1YE5WGx9Q5nyEGVxo44K+DvOzSJzYCxOl+GJ97bw3IpCzhyeyNNXnkZCbB8bpupDNOgDWX0lvH0XbPoXDD0LLnmuw+VrzU4XH20t4821RXywuZQmp4tRyfHMHp/OzLHpDHL04goVY+DADiusdnwAX38MzbVW73TwGW299pRTrMsz+kJdhTXMse1da133zKcDa9WHMfDlAlh8jzV0Ne1XkHd94H5gHUtzPXzxF1wf/w8hDVUscn2T3zVfSt64cYcDvr031xTxs4UbSekXxQvX5HFSss2G6HxEgz7QGWOtZFh8D0goXPwUjJl9zN2r6pp4+8u9LFxbxLrdVYQITMpKYtZp6Zyfk0JUeA/UKWmoblvut+MDqNptbU/IbBtnz5zUs+Poxlj1WN5/0BrHnv0CDP1Gz72er9SUwTs/gs1vWx+EM5+2huxsqHWI5l8rN/ED82+uD19CuLiQvB/A5HuOWRZk7e5Kbnp5DfVNTv542VimnHyC3/5sTIPeLioKrQnIolXWBORFT0LU8c+o3VFWw1tri1m4toiS6gbiI8O4KDeV2eMzmDAs4cSHdpzN1nK/Anevfc8XVs2UiDjIPNsakvHXcr+SdfDGD6Byp3UCz6Qf990iXJvftr6xNR6Ecx+AM2/vu23thvKaRp5fUcjfP9t55BBNxEHrMpxrX4bwaPjGHXDmbR12CPZW1zP372v4qqSan54/mpvPHt43hib7CA16O3G2wIonYcUT0G+wNZSTkmuVk21pgOaGttuH79fjam6gsKScDTv3s6OkjBBnI0nRhjGDIjgpMYL40BaPY+qtpYst7t+e91uf3zjb2pQ61mO538QuLaPrMY2H4D93w8YF1uqdWX8BR6q/W9WmvgrevRe+fA1ST7X+Hged7O9W+dxRAT82ndvPHcnw9ssmy7dbSzI3L7Lmbs6+16qe2W7paH2Tk5+++SVvbyhh5tg0Hp99Ss98Qw1AGvR2tHslLLyxbYjkBLgIod6E00g4rtAoIqJiiY2NJTQiGsKi2n7CoyAs2loFE+7+HRYNCcNOrGZKbzEG1r8Ki39itfuS5/pGEbmCD+Dft1uT7JPvgck/6Rsfjj7kdcC3V7TaWpK582OIT4WcWdZZ0Bl5h+crjDHMW17A797fxqkZ/Xj++3kk9+YcVB+lQW9XDQet4lKu5nZB7BnQxwnskDBKqht4a10xb64torCslsiwEKblpDDrtHQmjRxojzNxy7bBG9dZxb3OvB2mPOSfk4waa2DpL2D1X2HgKLjk2cPLZu3ihAPekzHWh+GqF6xhQWeTdcJczkwYM8v6BinC+/n7uOv19cRHhfH81XmcOrh/D72rwKBBrzpljGFDUTUL1xaxaEMJVXXNJMVHMnNsGrPHZzA6JcCrazbXw/sPWOGRdpq15n5AZu+9/q7P4F83W3V6zrzNGo8Pt0/xOp8EfEfqq6wzbL9aaJ1z4WqxJvhzLoExs9hihnDD39dQeqiRJy89hRlj033yfgKRBr3qkqYWF8u2lPLm2iKWbymlxWXITnUwe3wG009NIym+kxOZ+rJNi6x16sbAxX847uoln2hugOWPwqd/ti7mMfMZGHZWz75mL2oN+Jc/20Vji9N3Ad+RugrY8h8r9L9eYc0TJWZRd9J0HtiexcIiB7ecM4J7po0iJCT4Jmk16NUJq6htYtH6YhauK+bLompCQ4RzTkpi1mkZfGt0gNbbqdxlnWBV9IW7zMTjEBHj+9cpWWed8Fa2BcZfB9Me7Z1qkr2gVwO+I7Xl1sTtVwth5yeAYX9UJv84NJ6KzO9w31XfIT7KXvMendGgVz6xbf8hFq4t5q11Rew/2IiIVV1zZFIcWcnxjEyKY2RyHCMHxeHo6//JnM2w/NfwyVPWBdsvfdF3F85wNlulGVY8CbFJMP3PkDXVN8/tZ34P+I4c2g+b/o3JXwi7VyIYCkIySTz9MhImzOndITo/0qBXPuV0GT7bcYA1uyopKKth+/5DFJbX0tTSdm34ZEckWYPiGTko7vBP1qA4EuP62LDPjmXWuQmNh+DC31o9/O6szS7dAm/dBHvXQ+73rDo10Qk+a66/tA/4Ge6A73PVJQ+WUPjRKxxa8zqnst3aljbOGtPPueToa+HaiAa96nFOl2FPRR0FpTVsL61he+khdpTWUFBaQ21T25r7hJhwsgbFM8Id/CMHxZGVHEeKI8p/J78c2m+Fc+FyKwwu/iNE9evac7ic8Nk8WOYenvnOU1alxgBXdqiRv3wcAAHfzq4DtfzsxcXkVi3nxgHrGXjQfUG8jAnWks3sGdDPXhO3GvTKb4wx7K1uYLs79AtKD7F9v/VhUF3ffHi/uMiwI8Pf/TsjIYbQ3phYc7ng//5gBXW/DGtVTkaH/2eOVlEI/7oNdn8Ko75tTfIGSCnhFqeLvdUN7KmoY09lHbsr6thTUc/uijqKKusor2kiRAiYgPd0qKGZO19bz7Itpdw+NpQfpeUTuumttqJ6Q85sC/0TLajXh2jQqz7HGEN5TdPh8G/7JlBD2aHGw/tFhoUwPOnID4ARg+LISIjumYngPV/AG9fDoRLrgtdn3nHswmtHlBMOhQufgFMv61OFyIwxVNQ2saeynj0VdYcDvDXQi6vqcbraMiAsREjrH82QATEMHhDN4AExnJ+TElAB78npMjyxZAvPfVTI6ZkDeOaq8Qyo3w35b0H+QijdBBJiFQzMucQK/b56AmAnNOhVQKmua6agzB3++2vc8wA1FFfVH7FfQkw46QnRpPePJr1/DGn9o8hIsG6nJ0STEBN+YsNB9ZXWpR03v22V05357NHFto4oJ3wOzJjnt3LC9U1Oj/CuY3dFPXsqrdt7KuqOGDoDGBgXQUZCzOEwt37HMDghhtR+UfY4Sa6dt9YVce+bG0l2RPLC9ycwKsVdS6d0ixX4Xy2EA9utooGZk2DYN63zLdLGBUwVVA16ZQu1jS0UltWyo8wK/eKqeoor237XNx8ZaNHhoaT1jyI9IYb0/tHuD4Fo0vpHk54QTXJ85LFDzRjrDNb37ofo/jDreSvQ25cTPu8RmHBDj/XijTHUNzutXnm7AN9dUceeyvojvgG1vu/WAG8L9Bj3/WhiIwNwSawPrHNXwKxtbOGpOWOZlpPS9qAxsD/fCv3N/4HyrW2PJWRaZzCnnWb9Tj21T14HV4Ne2Z4xhqq6Zoqr6inyCP+S1g+EqnoqapuOOCY0REhxRHl8K4huu+3+HXVgs1U+oXw7fPMuOFDQaTlhp8tQ29RCbWMLNQ0t1DS2UNvopKax9XbLUbfbtjmpbWw7trapBVe7/6IhAqn9oo/ukbt75QPjIrSq4zHsq25g7sur2VhczU+mjeLWc0Z0/GfVUA0l66FkLRSvgeJ1cLDIekxCIGm0O/jHWb+Tx/j92r0a9EoBdU0tlFQ1eHwTqLPuuz8Y9h1sOGK8GiAxNoLh/YW7ml7grEPv0iLhLE25kSWO2dQ0GY/Adh4O7Lp2QyXHEhYixEaGEef+iY0MJTYyjPioMGIjwtpuR4bRLzqcwQlWsKf1j9bL6nVDQ7OTn77xJYs2lDD91DSeuNTLCpg1pVC81h3+7t91B6zHQiOsKrKtvf6002BgVq+WnNagV8oLLU4X+w42uD8M6tqGhaoaKK6sY1jV55SYREojhxLnDuO4yDDrdmQYcZGh1rao1uB2h/Xh26EeoR5GZFiI9rz9xBjD0x/u4Hfvb2VMWj/+8v08Uvp1sQKmMVb12MPBv876aaqxHo+Iswqwtfb600+D/kN7bJhPg14ppTqwdNN+7nptHTGRYTx/9XjGDenmyW0upzXM59nr37fRqsAJEJNoTfCmj28Lfx8txdWgV0qpY9i67xA3/H0V+w828tvZuVwyzserp1qaoDTfY9hnHZRtBuM+k9yR0dbrHzzRWvFzAjTolVLqOCpqm7j1lTWsLKxg7uThXDo+g4SYCBJiwntmuWlTLezdcOSYf+XX1pm7N/z3hJ5Sg14ppTrR7HTx8KJ8Xvn8yKu2OaLCSIyLJCEmnAGxESTERFi/Y63fA2KOvO2IDjuxuZe6Cmtyd2DWCbVfg14ppby0fk8VRZV1VNY2caC2icraJirqmq3fHj9NTleHx4eGiPvDIPzwh0Lrj+eHRGLrh0VMBNER3V+dc7ygD84zJ5RS6hjGDu7P2E4uS2iMoa7JOpGtss7jA8F9v6K2mYraRiprm9leWkOle3v7cyJaRYWHMCAmgvHDBvCny8f5/D1p0CulVBeJyOHls4MHeHfRGpfLUF3fTEXdkR8KbR8SzaT065ky3hr0SinVC0JChAT3cA1Jne/v09f2ZicRuUBEtopIgYjc18Hjk0VkrYi0iMil7R67RkS2u3+u8VXDlVJKeafToBeRUGAecCGQDVwuIu2vubYbuBZ4td2xA4CHgNOBicBDIhL4l9tRSqkA4k2PfiJQYIwpNMY0Aa8BR1w6xxiz0xjzJdB+Gvp8YKkxpsIYUwksBS7wQbuVUkp5yZugTwf2eNwvcm/zhlfHishcEVktIqvLysq8fGqllFLe6BMl8Iwxzxtj8owxeUlJvTxLoZRSNudN0BcDgz3uZ7i3eaM7xyqllPIBb4J+FZAlIpkiEgFcBizy8vmXANNEJME9CTvNvU0ppVQv6TTojTEtwO1YAb0ZWGCMyReRR0RkOoCITBCRIuC7wHMiku8+tgL4FdaHxSrgEfc2pZRSvaTP1boRkTJgVzeeYiBQ7qPmBIpge8/B9n5B33Ow6M57HmqM6XCSs88FfXeJyOpjFfaxq2B7z8H2fkHfc7DoqffcJ1bdKKWU6jka9EopZXN2DPrn/d0APwi29xxs7xf0PQeLHnnPthujV0opdSQ79uiVUkp50KBXSimbs03Qd1Yz325EZLCILBeRTSKSLyJ3+rtNvUVEQkVknYj8x99t6Q0i0l9E3hCRLSKyWUTO9HebepqI/Mj97/orEflfEYnyd5t8TUTmi0ipiHzlsW2AiCx1X79jqa/Kutsi6L2smW83LcCPjTHZwBnAbUHwnlvdiXWWdrD4I/CeMWY0cCo2f+8ikg78EMgzxowBQrFKr9jNSxxdtv0+4ANjTBbwgft+t9ki6PGiZr7dGGP2GmPWum8fwvrP72356IAlIhnAt4EX/N2W3iAi/YDJwF8BjDFNxpgqvzaqd4QB0SISBsQAJX5uj88ZY1YA7UvCzAD+5r79N2CmL17LLkHfnZr5AU9EhgHjgM/93JTe8Afgpxx9kRu7ygTKgBfdw1UviEisvxvVk4wxxcDvsK5ctxeoNsa8799W9ZpkY8xe9+19QLIvntQuQR+0RCQOeBO4yxhz0N/t6Uki8h2g1Bizxt9t6UVhwGnAM8aYcUAtPvo631e5x6VnYH3IpQGxInKVf1vV+4y19t0n69/tEvRBWfdeRMKxQv4VY8xCf7enF5wFTBeRnVjDc+eKyD/826QeVwQUGWNav629gRX8djYV+NoYU2aMaQYWAt/wc5t6y34RSQVw/y71xZPaJei7UzM/IImIYI3bbjbG/N7f7ekNxpifGWMyjDHDsP6OlxljbN3TM8bsA/aIyCj3pinAJj82qTfsBs4QkRj3v/Mp2HwC2sMi4Br37WuAf/viScN88ST+ZoxpEZHWmvmhwHxjTL6fm9XTzgKuBjaKyHr3tvuNMYv91yTVQ+4AXnF3YgqB6/zcnh5ljPlcRN4A1mKtLluHDcshiMj/AucAA93X83gIeBxYICLXY5Vr/55PXktLICillL3ZZehGKaXUMWjQK6WUzWnQK6WUzWnQK6WUzWnQK6WUzWnQK6WUzWnQK6WUzf1/etAsJY1MhX8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model3,loss3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "5895d641",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_25\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_50 (Conv1D)          (None, 118, 128)          12800     \n",
      "                                                                 \n",
      " conv1d_51 (Conv1D)          (None, 108, 64)           90176     \n",
      "                                                                 \n",
      " max_pooling1d_25 (MaxPoolin  (None, 54, 64)           0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_33 (Dropout)        (None, 54, 64)            0         \n",
      "                                                                 \n",
      " flatten_25 (Flatten)        (None, 3456)              0         \n",
      "                                                                 \n",
      " dense_50 (Dense)            (None, 125)               432125    \n",
      "                                                                 \n",
      " dense_51 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 535,857\n",
      "Trainable params: 535,857\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 15s 57ms/step - loss: 0.3541 - accuracy: 0.8538 - val_loss: 0.1622 - val_accuracy: 0.9279\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 16s 62ms/step - loss: 0.1566 - accuracy: 0.9361 - val_loss: 0.1289 - val_accuracy: 0.9426\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 16s 64ms/step - loss: 0.1434 - accuracy: 0.9434 - val_loss: 0.1599 - val_accuracy: 0.9353\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 17s 66ms/step - loss: 0.1221 - accuracy: 0.9492 - val_loss: 0.1416 - val_accuracy: 0.9462\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 17s 69ms/step - loss: 0.1194 - accuracy: 0.9485 - val_loss: 0.2834 - val_accuracy: 0.9131\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 18s 71ms/step - loss: 0.1435 - accuracy: 0.9444 - val_loss: 0.1231 - val_accuracy: 0.9494\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 18s 71ms/step - loss: 0.1173 - accuracy: 0.9497 - val_loss: 0.1104 - val_accuracy: 0.9513\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 19s 76ms/step - loss: 0.1112 - accuracy: 0.9537 - val_loss: 0.1129 - val_accuracy: 0.9497\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 18s 73ms/step - loss: 0.1114 - accuracy: 0.9525 - val_loss: 0.1102 - val_accuracy: 0.9500\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 21s 82ms/step - loss: 0.1284 - accuracy: 0.9500 - val_loss: 0.1233 - val_accuracy: 0.9455\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 18s 73ms/step - loss: 0.1303 - accuracy: 0.9518 - val_loss: 0.1225 - val_accuracy: 0.9487\n"
     ]
    }
   ],
   "source": [
    "# Next model: filter size = 64, kernel size = 11\n",
    "\n",
    "model4, loss4, prop4 = getModel(epochs_nr,batch_val, 128, 11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "2fd3cd0b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 13ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       532\n",
      "           1       1.00      1.00      1.00       462\n",
      "           2       1.00      1.00      1.00       421\n",
      "           3       0.84      0.87      0.85       530\n",
      "           4       0.88      0.84      0.86       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.95      3120\n",
      "   macro avg       0.95      0.95      0.95      3120\n",
      "weighted avg       0.95      0.95      0.95      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[532   0   0   0   0   0]\n",
      " [  0 462   0   0   0   0]\n",
      " [  0   0 421   0   0   0]\n",
      " [  0   0   0 462  68   0]\n",
      " [  0   1   0  91 494   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model4,loss4)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03928b7f",
   "metadata": {},
   "source": [
    "* the best properties for the kernel-size: 7, for lower is overfitted, larger is with higher loss value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "d062a338",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_23\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_46 (Conv1D)          (None, 122, 64)           4096      \n",
      "                                                                 \n",
      " conv1d_47 (Conv1D)          (None, 116, 32)           14368     \n",
      "                                                                 \n",
      " max_pooling1d_23 (MaxPoolin  (None, 58, 32)           0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_31 (Dropout)        (None, 58, 32)            0         \n",
      "                                                                 \n",
      " flatten_23 (Flatten)        (None, 1856)              0         \n",
      "                                                                 \n",
      " dense_46 (Dense)            (None, 125)               232125    \n",
      "                                                                 \n",
      " dense_47 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 251,345\n",
      "Trainable params: 251,345\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 6s 24ms/step - loss: 0.3940 - accuracy: 0.8353 - val_loss: 0.1677 - val_accuracy: 0.9260\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 6s 25ms/step - loss: 0.1610 - accuracy: 0.9309 - val_loss: 0.1366 - val_accuracy: 0.9462\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 6s 26ms/step - loss: 0.1457 - accuracy: 0.9412 - val_loss: 0.1334 - val_accuracy: 0.9446\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 7s 29ms/step - loss: 0.1260 - accuracy: 0.9468 - val_loss: 0.1268 - val_accuracy: 0.9462\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 7s 28ms/step - loss: 0.1222 - accuracy: 0.9492 - val_loss: 0.1215 - val_accuracy: 0.9462\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 7s 29ms/step - loss: 0.1145 - accuracy: 0.9510 - val_loss: 0.1188 - val_accuracy: 0.9474\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 7s 28ms/step - loss: 0.1158 - accuracy: 0.9527 - val_loss: 0.1267 - val_accuracy: 0.9497\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 8s 31ms/step - loss: 0.1243 - accuracy: 0.9493 - val_loss: 0.1173 - val_accuracy: 0.9513\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 7s 29ms/step - loss: 0.1118 - accuracy: 0.9526 - val_loss: 0.1139 - val_accuracy: 0.9471\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 7s 29ms/step - loss: 0.1085 - accuracy: 0.9559 - val_loss: 0.1147 - val_accuracy: 0.9481\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 7s 30ms/step - loss: 0.1124 - accuracy: 0.9522 - val_loss: 0.1187 - val_accuracy: 0.9490\n"
     ]
    }
   ],
   "source": [
    "# Next model: filter size = 64, kernel size = 7\n",
    "\n",
    "model5, loss5, prop5 = getModel(epochs_nr,batch_val, 64, 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "9815b1f9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 7ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.98      0.99       532\n",
      "           1       0.99      1.00      0.99       462\n",
      "           2       1.00      1.00      1.00       421\n",
      "           3       0.87      0.84      0.86       530\n",
      "           4       0.86      0.89      0.87       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.95      3120\n",
      "   macro avg       0.95      0.95      0.95      3120\n",
      "weighted avg       0.95      0.95      0.95      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[522   6   0   0   4   0]\n",
      " [  0 460   0   2   0   0]\n",
      " [  0   0 421   0   0   0]\n",
      " [  0   0   0 446  84   0]\n",
      " [  0   0   0  63 523   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model5,loss5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "f49819ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_22\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_44 (Conv1D)          (None, 122, 32)           2048      \n",
      "                                                                 \n",
      " conv1d_45 (Conv1D)          (None, 116, 16)           3600      \n",
      "                                                                 \n",
      " max_pooling1d_22 (MaxPoolin  (None, 58, 16)           0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_30 (Dropout)        (None, 58, 16)            0         \n",
      "                                                                 \n",
      " flatten_22 (Flatten)        (None, 928)               0         \n",
      "                                                                 \n",
      " dense_44 (Dense)            (None, 125)               116125    \n",
      "                                                                 \n",
      " dense_45 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 122,529\n",
      "Trainable params: 122,529\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 5s 16ms/step - loss: 0.4591 - accuracy: 0.8103 - val_loss: 0.2017 - val_accuracy: 0.9183\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 4s 17ms/step - loss: 0.1943 - accuracy: 0.9229 - val_loss: 0.1532 - val_accuracy: 0.9417\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 4s 17ms/step - loss: 0.1599 - accuracy: 0.9353 - val_loss: 0.1559 - val_accuracy: 0.9337\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 4s 18ms/step - loss: 0.1486 - accuracy: 0.9411 - val_loss: 0.1292 - val_accuracy: 0.9462\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 5s 18ms/step - loss: 0.1308 - accuracy: 0.9453 - val_loss: 0.1346 - val_accuracy: 0.9429\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 5s 19ms/step - loss: 0.1319 - accuracy: 0.9472 - val_loss: 0.1210 - val_accuracy: 0.9490\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 5s 19ms/step - loss: 0.1223 - accuracy: 0.9501 - val_loss: 0.1258 - val_accuracy: 0.9481\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 5s 19ms/step - loss: 0.1179 - accuracy: 0.9526 - val_loss: 0.1490 - val_accuracy: 0.9388\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 5s 19ms/step - loss: 0.1196 - accuracy: 0.9501 - val_loss: 0.1168 - val_accuracy: 0.9487\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 5s 19ms/step - loss: 0.1171 - accuracy: 0.9516 - val_loss: 0.1282 - val_accuracy: 0.9465\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 5s 20ms/step - loss: 0.1152 - accuracy: 0.9540 - val_loss: 0.1421 - val_accuracy: 0.9433\n"
     ]
    }
   ],
   "source": [
    "# Next model: filter size = 32, kernel size = 7\n",
    "\n",
    "model6, loss6, prop6 = getModel(epochs_nr,batch_val, 32, 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "12c1a417",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 0s 3ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.98      0.98       532\n",
      "           1       0.96      1.00      0.98       462\n",
      "           2       0.99      0.99      0.99       421\n",
      "           3       0.82      0.91      0.86       530\n",
      "           4       0.91      0.81      0.86       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.94      3120\n",
      "   macro avg       0.95      0.95      0.95      3120\n",
      "weighted avg       0.94      0.94      0.94      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[520   9   3   0   0   0]\n",
      " [  0 462   0   0   0   0]\n",
      " [  0   4 417   0   0   0]\n",
      " [  2   1   0 481  46   0]\n",
      " [  2   4   0 106 474   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model6,loss6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "bd067376",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_24\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv1d_48 (Conv1D)          (None, 122, 256)          16384     \n",
      "                                                                 \n",
      " conv1d_49 (Conv1D)          (None, 116, 128)          229504    \n",
      "                                                                 \n",
      " max_pooling1d_24 (MaxPoolin  (None, 58, 128)          0         \n",
      " g1D)                                                            \n",
      "                                                                 \n",
      " dropout_32 (Dropout)        (None, 58, 128)           0         \n",
      "                                                                 \n",
      " flatten_24 (Flatten)        (None, 7424)              0         \n",
      "                                                                 \n",
      " dense_48 (Dense)            (None, 125)               928125    \n",
      "                                                                 \n",
      " dense_49 (Dense)            (None, 6)                 756       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 1,174,769\n",
      "Trainable params: 1,174,769\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/11\n",
      "251/251 [==============================] - 26s 103ms/step - loss: 0.3302 - accuracy: 0.8683 - val_loss: 0.2499 - val_accuracy: 0.8881\n",
      "Epoch 2/11\n",
      "251/251 [==============================] - 28s 110ms/step - loss: 0.1506 - accuracy: 0.9404 - val_loss: 0.2050 - val_accuracy: 0.9250\n",
      "Epoch 3/11\n",
      "251/251 [==============================] - 25s 98ms/step - loss: 0.1301 - accuracy: 0.9453 - val_loss: 0.1454 - val_accuracy: 0.9410\n",
      "Epoch 4/11\n",
      "251/251 [==============================] - 25s 98ms/step - loss: 0.1301 - accuracy: 0.9477 - val_loss: 0.1156 - val_accuracy: 0.9490\n",
      "Epoch 5/11\n",
      "251/251 [==============================] - 24s 97ms/step - loss: 0.1175 - accuracy: 0.9485 - val_loss: 0.1120 - val_accuracy: 0.9490\n",
      "Epoch 6/11\n",
      "251/251 [==============================] - 24s 96ms/step - loss: 0.1394 - accuracy: 0.9452 - val_loss: 0.1337 - val_accuracy: 0.9426\n",
      "Epoch 7/11\n",
      "251/251 [==============================] - 24s 95ms/step - loss: 0.1247 - accuracy: 0.9490 - val_loss: 0.1184 - val_accuracy: 0.9490\n",
      "Epoch 8/11\n",
      "251/251 [==============================] - 24s 96ms/step - loss: 0.1127 - accuracy: 0.9532 - val_loss: 0.1170 - val_accuracy: 0.9516\n",
      "Epoch 9/11\n",
      "251/251 [==============================] - 24s 96ms/step - loss: 0.1099 - accuracy: 0.9504 - val_loss: 0.1069 - val_accuracy: 0.9500\n",
      "Epoch 10/11\n",
      "251/251 [==============================] - 24s 95ms/step - loss: 0.1083 - accuracy: 0.9522 - val_loss: 0.1418 - val_accuracy: 0.9365\n",
      "Epoch 11/11\n",
      "251/251 [==============================] - 24s 95ms/step - loss: 0.1219 - accuracy: 0.9508 - val_loss: 0.1123 - val_accuracy: 0.9490\n"
     ]
    }
   ],
   "source": [
    "# Next model: filter size = 256, kernel size = 7\n",
    "\n",
    "model7, loss7, prop7 = getModel(epochs_nr,batch_val, 256, 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "e95ea017",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 12ms/step\n",
      "--------------------------------Classification report-\n",
      " \n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       532\n",
      "           1       1.00      1.00      1.00       462\n",
      "           2       1.00      1.00      1.00       421\n",
      "           3       0.83      0.89      0.86       530\n",
      "           4       0.89      0.83      0.86       586\n",
      "           5       1.00      1.00      1.00       589\n",
      "\n",
      "    accuracy                           0.95      3120\n",
      "   macro avg       0.95      0.95      0.95      3120\n",
      "weighted avg       0.95      0.95      0.95      3120\n",
      "\n",
      " \n",
      "-------------------------------------Confusion matrix-\n",
      " \n",
      "[[532   0   0   0   0   0]\n",
      " [  1 461   0   0   0   0]\n",
      " [  0   0 421   0   0   0]\n",
      " [  0   0   0 472  58   0]\n",
      " [  0   1   0  99 486   0]\n",
      " [  0   0   0   0   0 589]]\n",
      " \n",
      "-------------------------------------Loss Values plot-\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Properties of the model visualisation\n",
    "propModel(model7,loss7)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1796f1e",
   "metadata": {},
   "source": [
    "# Conclusions\n",
    "* the best results achieved in the model3, loss values plot looks stable during epochs\n",
    "* accuracy in each model around 95%, the highest in model3\n",
    "* models sometimes hava problems to distinguish between activities number 5 and 6\n",
    "* probably the reason is that during the sitting and standing signals from sensors are similar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "4d583963",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as _jit_compiled_convolution_op, _jit_compiled_convolution_op while saving (showing 2 of 2). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: GeneralModel/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: GeneralModel/assets\n"
     ]
    }
   ],
   "source": [
    "# Saving the model\n",
    "model3.save('GeneralModel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "e95dc3eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "model3 = keras.models.load_model('GeneralModel');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "75d1c7c6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "98/98 [==============================] - 1s 6ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x1008 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# The confusion matrix visualisation\n",
    "y_pred = model3.predict(X_test)\n",
    "# Changing form of output data\n",
    "y_pred = np.argmax(y_pred, axis=1)\n",
    "y_real = np.argmax(y_test_cat, axis=1)\n",
    "# Counting the confusion matrix\n",
    "confusion_mtx = confusion_matrix(y_real,y_pred)\n",
    "# Labels \n",
    "LABELS = ['WALKING','WALKING_UPSTAIRS','WALKING_DOWNSTAIRS','SITTING','STANDING','LAYING']\n",
    "# Visualisation\n",
    "plt.figure(figsize=(16, 14))\n",
    "sns.set(style='whitegrid', palette='muted', font_scale=1.5)\n",
    "sns.heatmap(confusion_mtx, xticklabels=LABELS, yticklabels=LABELS, annot=True, fmt=\"d\", cmap=\"OrRd\");\n",
    "plt.title(\"Confusion matrix\", fontsize=25)\n",
    "plt.ylabel('True label', fontsize=25)\n",
    "plt.xlabel('Predicted label', fontsize=25)\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "a3faeecc",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-07-09 00:50:06.451277: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:362] Ignored output_format.\n",
      "2022-07-09 00:50:06.453872: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:365] Ignored drop_control_dependency.\n",
      "2022-07-09 00:50:06.484693: I tensorflow/cc/saved_model/reader.cc:43] Reading SavedModel from: GeneralModel\n",
      "2022-07-09 00:50:06.494749: I tensorflow/cc/saved_model/reader.cc:81] Reading meta graph with tags { serve }\n",
      "2022-07-09 00:50:06.495298: I tensorflow/cc/saved_model/reader.cc:122] Reading SavedModel debug info (if present) from: GeneralModel\n",
      "2022-07-09 00:50:06.514709: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:354] MLIR V1 optimization pass is not enabled\n",
      "2022-07-09 00:50:06.519098: I tensorflow/cc/saved_model/loader.cc:228] Restoring SavedModel bundle.\n",
      "2022-07-09 00:50:06.650022: I tensorflow/cc/saved_model/loader.cc:212] Running initialization op on SavedModel bundle at path: GeneralModel\n",
      "2022-07-09 00:50:06.668491: I tensorflow/cc/saved_model/loader.cc:301] SavedModel load for tags { serve }; Status: success: OK. Took 183815 microseconds.\n",
      "2022-07-09 00:50:06.781670: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:263] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "# Convert the model\n",
    "converter = tf.lite.TFLiteConverter.from_saved_model('GeneralModel') # path to the SavedModel directory\n",
    "tflite_model = converter.convert()\n",
    "\n",
    "# Save the model.\n",
    "with open('model.tflite', 'wb') as f:\n",
    "    f.write(tflite_model)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}